Virtually every company I speak to wishes they could do a more accurate job of forecasting sales revenues. The issue is often particularly challenging for B2B-focused companies with lengthy, complex sales process.
There’s no doubt that there is significant room for improvement. CSO Insights research suggests that less than 50% of forecasted deals close at the projected time and for the projected amount.
Given the variables involved, it’s hard to imagine that vendors with highly-complex, lengthy sales processes could regularly achieve 100% sales forecast accuracy – but many sales organisations are consistently doing far better than the CSO Insights figures would suggest.
How are they achieving this? There are a handful of common principles that serve to support a robust and accurate forecasting process:
- Consistent application of pipeline stage definitions and milestones
- Basing forecasting on a combination of fact and judgement
- Forecasting at both the opportunity and territory level
- Measuring sales people on sales forecast accuracy
Consistent pipeline stage definitions and milestones
You’ve got to start by publishing consistent pipeline stage definitions and milestones – and insisting that sales people conform to them. If this essential foundation isn’t properly laid, you have absolutely no chance of achieving acceptable sales forecast accuracy.
Recommended for YouWebcast: The Art of Growth Hacking: Gaining Early Traction by Doing Things that Don't Scale
Your chosen pipeline stages must be carefully chosen, clearly defined, simple to understand and they must reflect key steps in your prospect’s buying decision process and not just your sales people’s activity – and they must be embedded into your CRM system.
In particular, the milestones that separate one stage from the next must reflect observable evidence of progress in the prospect’s decision-making process – for example, agreeing an evaluation plan. Each milestone must be an issue of provable fact, and this evidence must be captured in the relevant opportunity record.
Base forecasting on a combination of fact and judgement
Placing individual opportunities at the appropriate stage of the prospect’s buying decision process provides the factual foundation for the forecast. This is why it’s so critical that you insist that your sales people keep the CRM system up to date.
In fact, I strongly recommend that you reject any forecast that does not accurately reflect the current status of the opportunity as recorded in your CRM system. Once the facts have been clearly established and proven, it’s time for judgement to be applied, and it’s equally dependent on a clear and consistently implemented understanding of definitions.
Forecast at both the opportunity and territory level
The likelihood of any individual opportunity closing in the current forecast period is an issue of judgement. I’ve found the following categories to be useful and effective:
- CERTAIN: It is expected that almost all of the opportunities in the “certain” category will close in the current forecast period, the exceptions being associated with some completely unpredictable change in circumstances. Over time, at least 90% of such opportunities should end up closing at the predicted value in the predicted period
- PROBABLE: It is expected that most of the opportunities in the “probable” category will close in the current forecast period. Over time, at least 75% of such opportunities should end up closing at the predicted value in the predicted period
- POSSIBLE: It is expected that some of the opportunities in the “possible” category will close in the current forecast period. Over time, at least 25% of such opportunities should end up closing at the predicted value in the predicted period
These judgements must be backed by what I’ve referred to as a “common sense interlock”. Unless opportunities have provably reached the appropriate advanced stage in the pipeline, they cannot be flagged as “CERTAIN”. Similar principles must apply to the PROBABLE and POSSIBLE categories. The idea here is to eliminate wishful thinking that is unsupported by either the facts or the past outcomes of similar opportunities.
In addition to accurately assigning probabilities at the individual opportunity level, sales people should also be expected to come up with accurate territory-level forecasts for the current forecast period. These aggregated forecasts must take into account all the individual opportunity-level forecast categories, but additionally require that the sales person apply an additional level of personal judgement and commitment.
I’ve found the following territory-level forecast definitions to be useful and effective:
- The COMMIT number is the minimum revenue value the sales person commits to deliver from their territory in the given forecast period. Sales people should be expected to rarely if ever deliver below their commit number
- The FORECAST number is the number the sales person actually expects to achieve. This number should be achieved at least 75% of the time, and over time the long -term achievement is expected to be at least 100%
Measure sales people on sales forecast accuracy
Last, but by no means least, sales people must be measured on their sales forecast accuracy – and a growing number of sales organisations have seen benefit from tying some element of the sales person’s compensation to their forecast accuracy.
They key thing when comparing forecast to actual is to ensure that you investigate the underlying causes of deviation, extract the lessons learned, and take targeted action to eliminate the sources of variation.
Sometimes, this will be connected to individual sales behaviour that is consistently either optimistic or conservative. But you’ll often find that the root cause lies in the sales process itself, and that a targeted programme or initiative can help to systematically improve performance.
How do these recommendations relate to your own experiences? Have you got other proven strategies you could share?